The present invention relates to an image pickup apparatus and, more particularly, to an image pickup apparatus for generating a video signal encoded to compress an information amount of a photographed image.
A digitalizing technique of a video signal is being improved and a high efficient coding of a moving image is being performed. As an example of the conventional technique of such a coding algorithm, there is an interframe predictive coding for coding a difference between frames in the video signal. Such a coding relates to a coding system for reducing an information amount by using a correlation between adjacent frame images. Further, there is known a motion compensation interframe predictive coding to improve a coding efficiency of an image including a fast motion by adding a motion compensation to the above system.
Such a coding system is adopted in an MPEG (Moving Picture Expert Group) as coding standard specifications for digital storage, H.261 of the ITU-T (International Telecommunication Union-Telecommunication Standardization Sector) as coding standard for communication such as a television telephone, and the like. As for a method for encoding/decoding in accordance with the above specifications, they are disclosed in, for example, the magazine of The Institute of Television Engineers of Japan, Vol. 45, No. 7, pages 793 to 799 and 807 to 812, 1991.
In those conventional methods, however, since the signal is encoded by using the correlation between the frames, there is a problem such that image quality is deteriorated in the cases where the present image is different from the preceding image or where the motion is fast, where the frame correlation is low and the motion compensation is difficult, for example, the case where the operation such as switching of the scene (scene change), fading for gradually erasing the image, zoom, or the like is performed. With respect to a technique for solving such a problem, there is, for example, a technique disclosed in the magazine of The Institute of Television Engineers of Japan, Vol. 35, No. 10, pages 949 to 954, 1985. However, according to the technique disclosed in such reference, a circuit scale increases because the fading prediction is performed. In particular, in the case where such an encoding function is installed in an image pickup apparatus in which miniaturization is required, an increase in circuit scale causes a large problem.